The turning point for Samantha Sutton came in the second year of her electrical engineering degree, while testing antilock car brakes. "There was something missing," she says. "I felt the average engineer in my division wasn't really hacking and constructing as I wanted to. They were fine-tuning, refining. I just didn't find it terribly exciting."

To satisfy her inner hacker, Sutton took an unusual step: she switched to biology after graduating. She is now building "circuits" from proteins rather than wires at the Massachusetts Institute of Technology, where she rubs shoulders with other geeks who get their kicks from taking microbes apart and putting them back together in novel ways. For them, genes and proteins are components to be plugged together at will. A bacterial cell, meanwhile, is regarded as a "chassis", to be stripped down and used as a circuit board and power supply.

They call themselves synthetic biologists, but their approach has little in common with traditional biology. Where a biologist sees a cell as a system to be understood in all its exquisite detail, Sutton and her colleagues want to strip away the complexity and create simpler cells that do cool stuff. Many, like Sutton, are engineers by training, and they take their inspiration from the pioneering hackers who invented modern computing through a similar "hands on" approach.

In short, Sutton and her colleagues are bio-hackers, trying to launch a revolution that some believe will have as big an economic impact as the computer industry. "I'm not very good at building biological devices right now, but I damned well know that I'm going to be," says Tom Knight, a computer scientist at MIT who nowadays spends most of his time tinkering with genes and cells. "In my opinion, this is the technology that is going to drive the coming century."

When Knight draws parallels with computing, people listen. In the 1960s, he was one of the obsessive hackers at MIT who coalesced around the first interactively programmable computers. This fabled group wrote operating systems and programs, built peripheral devices and laid the foundations for the computing revolution. "We rode that exponential, but in a lot of respects, the fun is out of that now," says Knight.

Today, Knight and his followers get their kicks hacking away at the genomes of various microbes, while dreaming of a future in which biological devices are as ubiquitous as today's electronics. It's an exciting vision, but also a little frightening. The original hackers were idealists, who believed that open access to computers could only make the world a better place. Soon, though, others turned the same skills to malign ends, writing malicious code and committing acts of cyber-vandalism. Already, it is possible to construct deadly viruses by assembling their DNA from scratch, and synthetic biologists are busy thrashing out a code of conduct to minimise any additional dangers their field might pose (see "What about bad guys?").

For now at least, there is nothing to be very scared about. When the journal Nature highlighted synthetic biology in a special issue last November, it published a paper describing how E. coli bacteria had been genetically engineered to behave as a living photographic film. Researchers led by Chris Voigt of the University of California, San Francisco, designed a simple circuit consisting of genes and proteins, including parts of a light receptor from a cyanobacterium. The team got a gene synthesis company to make the extra DNA needed, and spliced it into E. coli. The modified bacteria produce a black pigment in the presence of red light, allowing them to be used as an old-fashioned but very high-definition photographic film, albeit one that takes hours rather than milliseconds to respond to light.

This achievement places the bio-hackers roughly on a par with the Homebrew Computer Club of the 1970s, a group of enthusiasts based around California's Silicon Valley that got together to share hacking tips and tricks. Voigt's living film may do little for photographers, but then music lovers would scarcely have been impressed by the faltering melody that rasped from a radio owned by a former building contractor called Steve Dompier on 16 April 1975. For Homebrew's assembled geeks, however, this wheezing rendition of the Beatles' Fool on the Hill was a revelation. The radio was controlled by Dompier's kit-built computer, and the music confirmed that the machine could do more interesting things than simply flashing its array of LEDs to give the answer to a mathematical problem. The audience included such legends as Steve "Woz" Wozniak, who two years later unveiled the Apple II, the first computer designed for the masses. Today, the group's members are credited with launching a personal computing industry that has transformed our working lives and leisure time.

In one key respect, the Homebrew crew had a head start on today's bio-hackers. "When Apple got started, there was already lots of commodity electronics," says Drew Endy, a synthetic biologist at MIT. "Woz didn't have to build his own power supply from stuff he dug up in the hills." But that is still the sorry state of the art in commercial genetic engineering, says Endy. "It has a history of taking random bits of genetic material lying around in nature and reusing them, directly."

Making biobricks

Indeed, for bio-hackers, the term "genetic engineering" is a misnomer. First, the biotech industry rarely attempts anything much more sophisticated than getting E. coli to make large quantities of a single protein from another organism. Even then, it often takes extensive research to discover why a sequence borrowed in this way fails to work well out of its usual context. True engineers, notes Endy, build much more sophisticated systems, which slot together according to a predetermined design and work reliably. "Genetic engineering doesn't look or feel like any form of engineering," he says.

This is why one of the initial goals of MIT's synthetic biologists is to produce a library of standard components, or biobricks, that will perform reliably. At the simplest level, these are short stretches of DNA, including gene sequences that code for proteins and promoter sequences that control the activity of genes. The library also includes composite parts built from basic biobricks, forming simple genetic devices.

So far, MIT's Registry of Standard Biological Parts contains 167 basic biobricks, and 421 composite parts. There is a long way to go before it rivals electronics equivalents such as the TTL Catalog, which documents many thousands of components. But the project gets a significant boost each year from a competition in which students vie to build the coolest biological devices, manufacturing new biobricks along the way. First run for MIT students in 2003, this has now evolved into the annual International Genetically Engineered Machine competition. It is partly designed to foster an environment conducive to hacking, like the Homebrew Computing Club, and it seems to work: Voigt's device arose from his students' entry in 2004. The 2006 event will be the biggest yet, with 39 teams from around the world.

As well as standardising genetic components, synthetic biologists are trying to tame the cells in which their devices must be placed. "Let's get rid of all the dangling wires that might short out our circuits," says George Church, a geneticist and synthetic biologist at Harvard University. At the University of Wisconsin-Madison, geneticist Fred Blattner has made a good start in this endeavour, creating a stripped-down strain of E. coli, shorn of about 15 per cent of its genome.

Knight thinks that biologists are too attached to E. coli. Even Blattner's strain is a complex cell containing around 3700 genes, many of which might interfere with added circuits. So Knight's preferred chassis is a humble microbe called Mesoplasma florum, isolated from a flower in Florida, which has just 682 genes. Now he is working to create an even simpler version. "What I'm hoping to do is rewrite the genome of this organism," he says. "We're going to get rid of a lot of the junk and restructure what is left."

Endy has already demonstrated the principle using a simple virus known as the T7 phage, which infects E. coli. Not only did he remove DNA with an unknown function, but he also separated out the phage's overlapping genes. Endy describes the process as "refactoring", a software-engineering term for cleaning up messy code to make it easier to work with without changing its function.

Conventional biologists do not care much about parallels with software development, but they can at least appreciate that Endy's project has shown overlapping genes aren't necessary for the T7 phage to function. When he and Knight start talking about biobricks, however, most biologists simply glaze over. "They are not excited," says Knight. "Nor should they be. It's a different agenda."

Powerful tool

Those biologists who do grasp the engineering agenda of Endy and Knight tend to view the entire enterprise as doomed, because of the complexity and interdependence of genetic networks that have evolved over billions of years. That was certainly the experience of Jim Collins, a physicist at Boston University, who in 2000 created a device considered as one of the first forays into synthetic biology. His idea was to create a simple toggle switch, engineered into E. coli, based on two genes that try to turn one another off. "We talked to molecular biologists. They said don't bother, it's not going to work," Collins says.

Collins did not listen, and after a little tweaking, his switch worked just fine. His switch stays in one state, with one gene suppressing its partner, until an external signal temporarily disables the active gene, stably switching the device into the opposite state. Since then he has shown that his switch can perform useful functions when hooked up to various pathways, such as selectively shutting down particular genes. It can also determine whether a bacterial culture has reached a critical density, by detecting "quorum sensing" signals through which the cells communicate about their numbers, before turning on a particular gene.

Based on his experience, Collins believes the MIT vision of plug-and-play biobricks may be hard to achieve. But if the biobricks are well-characterised, the tweaking necessary to get them up and running in each new device should not be too onerous, he reckons. The naysayers are ignoring a powerful tool the pioneers of computing were not able to call upon: evolution. Once a genetic circuit is more-or-less functioning, a powerful way of optimising it is to randomly mutate any DNA sequences involved and select for variants that perform better. Researchers led by Frances Arnold of the California Institute of Technology in Pasadena have already demonstrated the principle with a genetic "logic gate" similar to the toggle switch.

Collins's switch is now being licensed to biotech companies, while a modified version of Voigt's "camera" could perhaps be used to lay down specific patterns of biological materials just as light is used to etch circuits on silicon chips. Admittedly, these projects are unlikely to generate the same level of excitement as the Apple II computer. Nor are they "killer apps" like the first word-processing and spreadsheet programs. It's difficult, at least at this early stage, to imagine synthetic biology ever pervading our lives to the same extent as personal computing. But thanks to another link-up with the world of computing, synthetic biology's first world-changing application may not be far off.

Bill Gates is now pouring millions of dollars into synthetic biology as part of his quest to defeat malaria. In December 2004, the Bill and Melinda Gates Foundation gave a whopping grant of $42.6 million to Jay Keasling, a chemical engineer at the University of California, Berkeley, who came up with an audacious plan to redesign and transfer to E. coli and yeast the complex biosynthetic pathway that produces the potent anti-malarial drug artemisinin in the Chinese shrub Artemisia annua. The goal is to save millions of lives by slashing the price of this relatively expensive drug.

Uncharted territory

This meant designing a circuit involving multiple genes and proteins, most of which had to be re-engineered to precisely control levels of activity. "It's a big project, and a daunting one," says Keasling. The pressure was on, as Gates gave him just three years to perfect the basic circuit. Last month in Nature, Keasling revealed that he was ahead of schedule, describing a strain of yeast with a circuit of eight additional genes that makes artemisinic acid, just one chemical step away from the final drug. This puts the project on target to scale up the technique for production by the end of 2009.

Keasling has applied synthetic biology's hacker ethic to a mammoth problem of conventional genetic engineering, but others foresee applications that move into uncharted territory, including computing and other areas today dominated by electronics. "I can imagine some kind of marriage between biology and inorganic engineering," says Church, who thinks a reasonable initial target would be to demonstrate simple part-biological devices that can make copies of themselves. "I think if somebody set that as a goal, we could do it in three years," he says.

Even the most enthusiastic bio-hackers find it hard to imagine where things might eventually lead - that is like expecting the pioneers of computing to have predicted the internet and the iPod. "I think it's just going to be mind-boggling," says Sutton. "The things we're thinking about now will seem so short-sighted."

 

What about bad guys?

Today, the term hacker is synonymous with criminals who vandalise computer systems. But it was coined as an accolade for a skilled programmer, and the original hackers are horrified by those who have given their calling a bad name by linking it with malicious attacks.

When it comes to synthetic biology, the stakes are even higher. Genetically engineered bioweapons are nothing new, but synthetic biology might provide a handy tool kit for anyone wanting to wreak havoc. "We are putting safety and public discussions as a top priority," says George Church, a geneticist and synthetic biologist at Harvard University.

This weekend, at the Synthetic Biology 2.0 meeting in Berkeley, California, researchers will be asked to sign up to a code of conduct designed to minimise the risks. So far, the main concern surrounds gene synthesis, by which genes of any desired sequence can be built from scratch. This means that the DNA of dangerous viruses or even bacteria could be ordered from companies that offer gene synthesis services - some of which perform few checks (New Scientist, 9 November 2005, p 8). The proposed code of conduct includes a resolution to stop placing orders with companies that do not follow best practice, and also calls on synthetic biologists to report colleagues planning dangerous experiments.

It is a good start, but in the long run, bio-hackers argue that their skills will also have to be deployed proactively to help design countermeasures against any kind of biological attack. That will be tough. But with gene synthesis and other techniques for engineering deadly pathogens already widely available, there may be no option but to try.

"There are two ways of dealing with dangerous technologies," says Tom Knight, a leading figure in synthetic biology at the Massachusetts Institute of Technology. "One is to keep the technology secret. The other one is do it faster and better than everyone else. My view is that we have absolutely no choice but to do the latter."

 

History repeats itself

Leading biotechnology companies such as Genentech like to regard themselves as cutting edge. And compared with conventional drugs companies, perhaps they are. But from another perspective, they are dinosaurs, the equivalent of 1960s' IBM, once reviled by hackers as the world's dullest computing company.

Historians of computing agree that IBM missed a trick by persisting for too long with its narrow view of computers as machines for doing business and research, and failing to see the potential of personal computing. To synthetic biologists, Genentech and its ilk, while skilled at splicing individual genes into bacteria, are largely blind to the wider possibilities.

"Biotech industries will all have to change immensely if this works," predicts Randy Rettberg, who helped write the communications protocols that underpin the internet. Today, he is at the Massachusetts Institute of Technology, helping his old hacker friend Tom Knight launch the field of synthetic biology.

So far few companies have embraced the ideas of synthetic biology. But some of the field's leading lights have formed their own company, called Codon Devices, which intends to design genetic circuits to produce drugs, biosensors and much more. And Craig Venter, whose previous company, Celera, attempted to sequence the human genome ahead of the publicly funded project, has created a stir by launching a company called Synthetic Genomics. It aims to design and build microbes for various purposes, such as producing biofuels.